Brightline Ridership Number-Crunching and Analysis

[cbustrains]

April 2019 results are out and they don't look pretty:

"For the month ended April 30, 2019, we carried 71,308 passengers and generated total revenue of approximately $1.8 million. Daily trips by commuters and frequent business travelers continued to grow in the month of April, while leisure-oriented ridership was lower due to lower seasonal event activity. In addition, we ran a reduced train schedule to allow for installation of signaling system upgrades associated with federally-mandated positive train control."

https://emma.msrb.org/ES1269861-ES993909-ES1395346.pdf

Does anyone have an idea of how much of a reduced train schedule occurred in April?

It seems that Virgin Trains USA is consistently releasing the monthly reports around the 18-20th of the following month. They also released a Construction Report stating that construction has commenced on the North Segment (West Palm Beach to Orlando) and will begin revenue service on that line in "approximately three years."

 

[chrsjrcj]

I honestly can say I’m not surprised. They started running a couple flash sales during April, and I even received a sales pitch for considerable discounts on their commuter passes.

They ran another flash sale today for up to 55% off in Smart for travel later this week.

As far as the reduced schedule, I seem to remember a Sunday or two where trains didn’t run all the way to Miami due to PTC installation.

I suspect May will rebound because of the $5 tickets this past weekend. There were a few sold out trains according to the booking site. Of course, that doesn’t look good from a revenue perspective, but I suppose they’re trying to get butts in the seats to sell them on the product first.

 

[Anderson]

Updated monthly ridership data:

      2018    2019   2020
Jan  17,783  73,568
Feb  24,098  78,707
Mar  32,899  91,903
Apr  ??????  71,308
May  ??????  85,740
Jun  ??????
Jul        
Aug        
Sep        
Oct  60,013
Nov  80,660
Dec  98,076

Revenue:
     2018    2019    2020
Jan          $1.7m
Feb          $1.9m
Mar          $2.3m
Apr          $1.8m
May          $1.7m
Jun         
Jul         
Aug         
Sep         
Oct  $1.0m   
Nov  $1.5m   
Dec  $2.2m

 

[Anderson]

      2018    2019   2020
Jan  17,783  73,568
Feb  24,098  78,707
Mar  32,899  91,903
Apr  ??????  71,308
May  ??????  85,740
Jun  ??????  80,094
Jul       
Aug       
Sep       
Oct  60,013
Nov  80,660
Dec  98,076
Revenue:
     2018    2019    2020
Jan          $1.7m
Feb          $1.9m
Mar          $2.3m
Apr          $1.8m
May          $1.7m
Jun          $1.6m
Jul         
Aug         
Sep         
Oct  $1.0m   
Nov  $1.5m   
Dec  $2.2m

 

[neroden]

Sooooo, the thing is, running a system like this is mostly fixed costs. The line as it is, by itself, is never going to break even. Going to Orlando should more-than-double revenue, while much-less-than-doubling costs. The same will happen when they go to Tampa.

Railroads of the 19th century expanded as fast as they possibly could for this reason. This also explains the merger mania tendency in railroads. Brightline's situation is no exception. Here's hoping they get to Tampa before they run out of funding. Then we can see how viable they are.

 

[Trogdor]

Railroads of the 19th century expanded as fast as they possibly could for this reason.

Railroads of the 19th century expanded as fast as they could because they were chasing government grants, an essentially non-existent market, and were defrauding lots of investors. Many of them ultimately went bankrupt.

 

[Anderson]

I think the answer is both. You had a lot of reckless expansion early on chasing government grants and charters...but scale advantages drove the various waves of mergers (and are part of why you saw "domino effect" waves of mergers where one tie-up would trigger several more).

 

[neroden]

There weren't any government grants for the Northeastern railroads -- they still all expanded and merged as fast as possible. They also were big on through traffic, connections, and "codesharing". Economies of scale were the only way to survive, and this was understood quite early.

 

[Anderson]

No, but there were definitely cases of subsidies and support for charters and the like. At least some mergers were, IIRC, driven by railroads trying to patch together the right to X miles of track for a given project.

 

[Anderson]

Updated for July data:

      2018    2019   2020
Jan  17,783  73,568
Feb  24,098  78,707
Mar  32,899  91,903
Apr  34,615  71,308
May  56,781  85,740
Jun  48,619  80,094
Jul  52,162  83,741  
Aug      
Sep      
Oct  60,013
Nov  80,660
Dec  98,076

NB: April/May 2018 ridership derived from April/May 2019 ridership adjusted for growth. The monthly totals do not quite line up with the YTD numbers for 2018 and I do not see any way to reconcile the discrepancy.

Revenue:
     2018    2019    2020
Jan          $1.7m
Feb          $1.9m
Mar          $2.3m
Apr          $1.8m
May          $1.7m
Jun  $0.8m   $1.6m  
Jul  $1.0m   $1.7m  
Aug      
Sep      
Oct  $1.0m
Nov  $1.5m
Dec  $2.2m

 

[Anderson]

     2018    2019   2020
Jan  17,783  73,568
Feb  24,098  78,707
Mar  32,899  91,903
Apr  34,615  71,308
May  56,781  85,740
Jun  48,619  80,094
Jul  52,162  83,741
Aug  54,574  74,312
Sep    
Oct  60,013
Nov  80,660
Dec  98,076

Revenue:
    2018    2019    2020
Jan          $1.7m
Feb          $1.9m
Mar          $2.3m
Apr          $1.8m
May          $1.7m
Jun  $0.8m   $1.6m
Jul  $1.0m   $1.7m
Aug  $0.9m   $1.5m 
Sep    
Oct  $1.0m
Nov  $1.5m
Dec  $2.2m

 

[neroden]

Perfectly respectable numbers, but they'll need more economies of scale to break even. Bring on Orlando.

 

[chrsjrcj]

I think they may fall just short of hitting 1 million riders this year.

You can see why they are rushing to add 3 additional stations to the current route.

 

[Anderson]

By "three stations"...I know there's Boca Raton, Aventura (I think)...and did they actually announce Fort Lauderdale Airport?

YTD they're averaging just over 79,900/month (or just over 2650/day if you go that way), putting them "on pace" to finish with about 960-970k (depending on if you go by "share of months elapsed" or "share of days elapsed"). FWIW, I would put the even-money point for them at one million: They probably lost around 5-10k riders over the hurricane (September will probably end up somewhere in the high 60s or low 70s) but based on last year, December is probably good for an extra 20k and November for an extra 5k.

(It also isn't quite clear how they handle riders on the Polar Express-ish trains in terms of headcount...or, for that matter, whether the month/year ends with 2359 on New Years' Eve or whether those late-night extras get handled as part of the "operational day" of 12/31 vs 01/01.)

 

[chrsjrcj]

PortMiami is the 3rd station. I had to look it up, because I too forgot what the 3rd one even was.

 

[jis]

Port Miami will only be for cruise related service. Hourly service there seems both impractical and unnecessary.

 

[Anderson]

PortMiami will likely be basically a "charter" service in conjunction with cruises. However, I feel like they're probably one set short of being able to really do that comfortably.

 

[Anderson]

September has been posted:
https://emma.msrb.org/ES1314134-ES1026580-ES1428550.pdf

     2018    2019   2020
Jan  17,783  73,568
Feb  24,098  78,707
Mar  32,899  91,903
Apr  34,615  71,308
May  56,781  85,740
Jun  48,619  80,094
Jul  52,162  83,741
Aug  54,574  74,312
Sep  52,850  61,688
Oct  60,013
Nov  80,660
Dec  98,076

    2018    2019    2020
Jan          $1.7m
Feb          $1.9m
Mar          $2.3m
Apr          $1.8m
May          $1.7m
Jun  $0.8m   $1.6m
Jul  $1.0m   $1.7m
Aug  $0.9m   $1.5m 
Sep  $1.0m   $1.1m
Oct  $1.0m
Nov  $1.5m
Dec  $2.2m

 

[neroden]

20% YOY growth is respectable. Still need Orlando, and possibly Tampa, to cover fixed costs.

 

[Anderson]

https://emma.msrb.org/ER1272506-ER993327-ER1396165.pdf

     2018    2019   2020
Jan  17,783  73,568
Feb  24,098  78,707
Mar  32,899  91,903
Apr  34,615  71,308
May  56,781  85,740
Jun  48,619  80,094
Jul  52,162  83,741
Aug  54,574  74,312
Sep  52,850  61,688
Oct  60,013  83,426
Nov  80,660
Dec  98,076

    2018    2019    2020
Jan          $1.7m
Feb          $1.9m
Mar          $2.3m
Apr          $1.8m
May          $1.7m
Jun  $0.8m   $1.6m
Jul  $1.0m   $1.7m
Aug  $0.9m   $1.5m 
Sep  $1.0m   $1.1m
Oct  $1.0m   $1.7m
Nov  $1.5m
Dec  $2.2m

 

[Brian_tampa]

https://emma.msrb.org/ER1272506-ER993327-ER1396165.pdf

     2018    2019   2020
Jan  17,783  73,568
Feb  24,098  78,707
Mar  32,899  91,903
Apr  34,615  71,308
May  56,781  85,740
Jun  48,619  80,094
Jul  52,162  83,741
Aug  54,574  74,312
Sep  52,850  61,688
Oct  60,013  83,426
Nov  80,660
Dec  98,076

    2018    2019    2020
Jan          $1.7m
Feb          $1.9m
Mar          $2.3m
Apr          $1.8m
May          $1.7m
Jun  $0.8m   $1.6m
Jul  $1.0m   $1.7m
Aug  $0.9m   $1.5m
Sep  $1.0m   $1.1m
Oct  $1.0m   $1.7m
Nov  $1.5m
Dec  $2.2m

There definitely seems to be a repeating pattern here. Late summer/early fall drop in ridership two years in a row. I wonder if the very good Oct '19 numbers portend even better numbers for Nov and Dec?

 

[Anderson]

I have several thoughts here:
(1) September is the "slow season" for Florida: It isn't summertime (with families traveling) and it isn't winter (with snowbirds).
(2) 2019 had an explicit disruption due to weather around Labor Day.
(3) On the other hand, I forget when hourly service was instituted last year (I know it was in place by mid-September but I forget if it was in mid-August or a little later), but the staggered service increases across much of last year do complicate comparisons.

Looking to the end of CY19, I think we're going to end up at around $16.5m in ticket revenue and $5.0m in ancillary revenue (so about $21.5m overall). Ridership should end up pretty close to 1,000,000: Though I think it's going to fall a hair shy, we seem to be holding at around 20,000 pax/month more than last year (aside from September, where they anticipate that 11k riders were lost due to Dorian) which could put us just over the line. To my mind, the bigger question is whether ridership in January holds to such an uptrend. On that, my best guess would be that YOY ridership will bump by about 10-15k/month...but we still don't have a good trend cycle to point to here.

 

[Anderson]

https://emma.msrb.org/ES1333216-ES1040101-ES1443376.pdf

     2018     2019    2020
Jan  17,783   73,568 
Feb  24,098   78,707
Mar  32,899   91,903
Apr  34,615   71,308
May  56,781   85,740
Jun  48,619   80,094
Jul  52,162   83,741
Aug  54,574   74,312
Sep  52,850   61,688
Oct  60,013   83,426
Nov  80,660  100,627
Dec  98,076

Note that Brightline crossed 100k pax for the first time in November. A trend of an increase of 20k riders/month seems to be holding. YTD ridership is 885,114. I'd give them a 40% chance of making it to one million riders in CY19.

    2018    2019    2020
Jan          $1.7m
Feb          $1.9m
Mar          $2.3m
Apr          $1.8m
May          $1.7m
Jun  $0.8m   $1.6m
Jul  $1.0m   $1.7m
Aug  $0.9m   $1.5m 
Sep  $1.0m   $1.1m
Oct  $1.0m   $1.7m
Nov  $1.5m   $2.2m
Dec  $2.2m

Again, overall revenue is looking strong. Average fares are up by about 20% per the monthly report and overall revenue is up 50% or so YOY. YTD revenue is $19.2m. Overall revenue looks likely to settle at about $22m.

 

[Anderson]

Just posting this as a thought, but I decided to check Brightline's numbers for the equivalent to Amtrak's FY19. Ridership for FY19 would be 939,810. It takes a little bit of work to sort out the other three months (Oct-Dec 2018), but ticket revenue for the nine months ended in September is indicated as $12.0m. For November 2018, it was $1.2m and for October 2018 it was $0.8m. I do not have December...but as October and November showed ticket revenue at about 80% of overall revenue (or put a different way, ancillary revenue is about 20% of the total), I'm comfortable putting $1.8m in as a placeholder (which is, incidentally, how the figures broke down for November 2019). That would give a total revenue of $15.8m for the period, probably +/- $0.1m.

The closest analogue in the Amtrak system would be the Hiawathas, though the Capitol Corridor and the Surfliners are also not utterly dissimilar (ditto the Springfield Shuttles, Downeaster, and NYP-ALB). When completed, the run to Orlando will probably look most like either the Surfliner or the Capitol Corridor in terms of distance and frequency.

I am partially noting this because I will likely, in the next few years, be adding a new page to my ridership/revenue data sets for non-Amtrak intercity operators and I want to keep an eye on the relevant data trail. Right now, non-Amtrak intercity ridership is about 2.8% of the total, but Brightline made up somewhere in the ballpark of 35% of the growth in intercity rail ridership in FY19.

I would also suspect that Brightline probably gave non-Amtrak intercity ridership its highest year in the US this side of the mid-70s, though I wouldn't be shocked if in '71 or '72 the four Amtrak holdouts could have scraped together the better part of a million riders (particularly since Southern was still running three daily trains plus a 3x/week run at the time). My guess is that about 65-70% of the ridership would have been Southern, with the Rio Grande and Rock Island splitting the remainder (the Rock Island was running two daily trains while the Rio Grande was running one 3x/week train...but it was quite popular with tourists) and the remaining operation in Georgia only picking up a few stray hundred a year based on anecdotal evidence. Brightline still probably beats the batch of them, however.

(As an observational aside, if Brightline, Texas Central, and the Vegas operation all get going to a full build-out there's every chance that private intercity train counts will quickly surpass where they were in 1971. As another aside, I've now blundered into a BTS file giving Amtrak's monthly ridership from 1991 to present, albeit with no breakdown by route or anything...)

 

[neroden]

If you're doing retro intercity calculations for 1971-1975, I'll remind you that the South Shore Line was considered intercity in 1971, and quite likely had higher ridership than the Rio Grande or the Rock Island, perhaps higher than the Southern, certainly higher than the Central of Georgia. The South Shore continued in fully private, unsubsidized operations through 1975, before getting state subsidy in 1977.

There were *six* railroads eligible to join Amtrak which didn't join Amtrak; the sixth was the Reading, with its Philadelphia-Newark service (the Crusader), which continued running through Reading's bankruptcy. I'm not sure when the start of SEPTA sponsorship was, but they also kept running to Bethlehem and Allenton and to Pottsville and Reading, which are definitely intercity services by any reasonable definition. Of course, this went into public ownership with Conrail and the 4R Act in 1976, and service ended in 1982.

The "intercity"/"commuter" distinction is pretty arbitrary, and Virgin Trains USA is currently running at the outer edge of commuter-distance operations, with a route shorter than the Port Jervis Line or the circuitous North Jersey Coast Line, and a bit longer than Metra's UP-NW line or the South Shore Line.

There's some difficulty creating apples-to-apples comparisons as a result. It'll be more clearly intercity once it gets to Orlando. It'll be a wildly different market from Miami-WPB, and it should be interesting to see what happens to the ridership numbers.